Fast and efficient feature importance calculations for EVOO spectroscopic data
Author(s)
Date Issued
2024
Type
article
Volume
47
Start Page
298
End Page
301
Journal
Abstract
Chemometrics via artificial neural networks is nowadays the state
of the art for food analysis. Here we consider an application aiming to distinguish
European (EU) extra vergin olive oil (EVOO) from non EU one. Though the interpretability
of neural network is still an open issue, here we propose a simple and
transparent method to “open the box” and showing the relative importance of the
spectral wavelengths in determining the classification output.